ValueError in unstructured Library: Mutable Default for numpy.ndarray Field in PDF Parsing

[ecentric]

Issue:

It seems that the numpy.ndarray field in the unstructured_inference.inference.elements module is not allowed to have a mutable default value. The error suggests that a default_factory should be used instead.

Has anyone encountered this issue or found a solution for handling mutable default errors with dataclasses in the unstructured package?

Any guidance would be appreciated!

Library versions:

{'numpy': '1.26.4', 'unstructured': '0.15.13', 'unstructured-inference': '0.7.37', 'langchain': '0.3.1', 'opencv-python': '4.10.0.84', 'pandas': '2.2.3', 'scipy': '1.14.1'}

Detailed Error

I'm encountering an error while extracting text from a PDF (hierarchy.pdf) using the unstructured library.

Exception: ValueError  
Message: mutable default <class 'numpy.ndarray'> for field texts is not allowed: use default_factory  

Code i was running

`
from langchain_community.document_loaders import UnstructuredPDFLoader

def extract_using_unstructured(file_path):
"""
Extract and process text from a PDF file using Unstructured via LangChain.

Args:
    file_path (str): The path to the PDF file.
    
Returns:
    str: The extracted text content from the PDF, or an error message if extraction fails.
"""
try:
    # Print version info for key libraries
    versions = {
        'numpy': importlib.metadata.version('numpy'),
        'unstructured': importlib.metadata.version('unstructured'),
        'unstructured-inference': importlib.metadata.version('unstructured-inference'),
        'langchain': importlib.metadata.version('langchain'),
        'opencv-python': importlib.metadata.version('opencv-python'),
        'pandas': importlib.metadata.version('pandas'),
        'scipy': importlib.metadata.version('scipy')
    }
    
    print(f"Library versions: {versions}")
    
    # Initialize the loader with the file path
    loader = UnstructuredPDFLoader(file_path)
    
    # Load the document
    docs = loader.load()

    # Extract text from all pages
    extracted_text = "\n".join([doc.page_content for doc in docs])
    return extracted_text

except Exception as e:
    # Get the full exception traceback
    tb = traceback.format_exc()
    
    # Log the detailed error
    print(f"Error extracting text from PDF {file_path}:\n"
                  f"Exception: {type(e).__name__}\n"
                  f"Message: {str(e)}\n"
                  f"Traceback:\n{tb}")
    
    # Return a detailed error message
    raise e`

Traceback:

Traceback (most recent call last): File "C:\Source\Utils\pdfparserstest.py", line 44, in extract_using_unstructured docs = loader.load() File "Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\langchain_core\document_loaders\base.py", line 31, in load return list(self.lazy_load()) File "Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\langchain_community\document_loaders\unstructured.py", line 107, in lazy_load elements = self._get_elements() File "C:\Users\sudie\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\langchain_community\document_loaders\pdf.py", line 72, in _get_elements from unstructured.partition.pdf import partition_pdf File "C:\Users\sudie\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\unstructured\partition\pdf.py", line 56, in <module> from unstructured.partition.pdf_image.analysis.layout_dump import ( File "C:\Users\sudie\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\unstructured\partition\pdf_image\analysis\layout_dump.py", line 7, in <module> from unstructured_inference.inference.elements import ImageTextRegion, TextRegion File "C:\Users\sudie\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\unstructured_inference\inference\elements.py", line 209, in <module> @dataclass File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.11_3.11.2544.0_x64__qbz5n2kfra8p0\Lib\dataclasses.py", line 1232, in dataclass return wrap(cls) File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.11_3.11.2544.0_x64__qbz5n2kfra8p0\Lib\dataclasses.py", line 1222, in wrap return _process_class(cls, init, repr, eq, order, unsafe_hash, File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.11_3.11.2544.0_x64__qbz5n2kfra8p0\Lib\dataclasses.py", line 958, in _process_class cls_fields.append(_get_field(cls, name, type, kw_only)) File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.11_3.11.2544.0_x64__qbz5n2kfra8p0\Lib\dataclasses.py", line 815, in _get_field raise ValueError(f'mutable default {type(f.default)} for field '